Wrap calls to dup() in shared_memory_posix.cc in HANDLE_EINTR.

Wrap calls to dup() in shared_memory_posix.cc in HANDLE_EINTR.

There's no reason that an interrupt to a dup() should cause a SharedMemory
operation to fail.

BUG=

Committed: https://crrev.com/fdbad9e261cffbd6cee672c5a8edbc5e58725531
Cr-Commit-Position: refs/heads/master@{#330584}

[erikchen, 2015-05-19 18:06:21]

erikchen@chromium.org changed reviewers:
+ rsesek@chromium.org

[erikchen, 2015-05-19 18:06:22]

rsesek: Please review.

Note that this CL was not prompted by an observed failures. I just ran across
this and it seemed reasonable to make the change.

[Robert Sesek, 2015-05-19 18:09:31]

LGTM. I wouldn't expect dup() to report EINTR (it's not really blocking), but
good catch.

[erikchen, 2015-05-19 18:12:48]

erikchen@chromium.org changed reviewers:
+ thestig@chromium.org

[erikchen, 2015-05-19 18:12:49]

thestig: Looking for an OWNER review.

[Lei Zhang, 2015-05-19 18:17:51]

lgtm

[erikchen, 2015-05-19 18:23:16]

The CQ bit was checked by erikchen@chromium.org

[commit-bot: I haz the power, 2015-05-19 18:24:31]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1138113006/1

[commit-bot: I haz the power, 2015-05-19 19:35:46]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2015-05-19 19:37:25]

Patchset 1 (id:??) landed as
https://crrev.com/fdbad9e261cffbd6cee672c5a8edbc5e58725531
Cr-Commit-Position: refs/heads/master@{#330584}